Skip to contentSkip to conversation

Check cfi_stubs before registering a struct_ops type.

PATCH Superseded [email protected] 3 participants 7 messages Feb 21, 2024 – Feb 22, 2024

Subsystems (5)

v4 of 5 Superseded Latest is v5 →

lore
Share & tools
lore
#0 lore raw

v4 cover Check cfi_stubs before registering a struct_ops type.

From: Kui-Feng Lee <[email protected]>

Recently, cfi_stubs were introduced. However, existing struct_ops types that are not in the upstream may not be aware of this, resulting in kernel crashes. By rejecting struct_ops types that do not provide cfi_stubs properly during registration, these crashes can be avoided.

---
Changes from v3:

- Remove CFI stub function for get_info.

- Allow passing NULL prog arg to check_member of struct bpf_struct_ops type.

- Call check_member to determines if a CFI stub function should be defined for an operator.

Changes from v2:

- Add a stub function for get_info of struct tcp_congestion_ops.

Changes from v1:

- Check *(void **)(cfi_stubs + moff) to make sure stub functions are provided for every operator.

- Add a test case to ensure that struct_ops rejects incomplete cfi_stub.

v3: https://lore.kernel.org/all/[email protected]/
v2: https://lore.kernel.org/all/[email protected]/
v1: https://lore.kernel.org/all/[email protected]/

Kui-Feng Lee (3):
bpf, net: allow passing NULL prog to check_member.
bpf: Check cfi_stubs before registering a struct_ops type. selftests/bpf: Test case for lacking CFI stub functions.

 kernel/bpf/bpf_struct_ops.c                   | 17 ++++
 net/bpf/bpf_dummy_struct_ops.c                |  2 +-
 tools/testing/selftests/bpf/Makefile          | 10 +-
 .../selftests/bpf/bpf_test_no_cfi/Makefile    | 19 ++++
 .../bpf/bpf_test_no_cfi/bpf_test_no_cfi.c     | 93 +++++++++++++++++++
 .../bpf/prog_tests/test_struct_ops_no_cfi.c   | 38 ++++++++
 tools/testing/selftests/bpf/testing_helpers.c |  4 +-
 tools/testing/selftests/bpf/testing_helpers.h |  2 +
 8 files changed, 181 insertions(+), 4 deletions(-)
 create mode 100644 tools/testing/selftests/bpf/bpf_test_no_cfi/Makefile
 create mode 100644 tools/testing/selftests/bpf/bpf_test_no_cfi/bpf_test_no_cfi.c
 create mode 100644 tools/testing/selftests/bpf/prog_tests/test_struct_ops_no_cfi.c
Signature · 1 line
2.34.1
#1 lore raw

PATCH v4 1/3 bpf, net: allow passing NULL prog to check_member.

From: Kui-Feng Lee <[email protected]>

To reuse the check_member function of the bpf_struct_ops structure for checking if an operator is supported, we permit passing a NULL value for the "prog" argument in check_member. The check_member function of the bpf_struct_ops structure will be utilized for checking cfi_stubs in a subsequent patch when registering a struct_ops type.

[email protected]

---
 net/bpf/bpf_dummy_struct_ops.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
net/bpf/bpf_dummy_struct_ops.c +1 -1
index 02de71719aed..5fe5461d3173 100644
@@ -178,7 +178,7 @@ static int bpf_dummy_ops_check_member(const struct btf_type *t,
 	case offsetof(struct bpf_dummy_ops, test_sleepable):
 		break;
 	default:
-		if (prog->aux->sleepable)
+		if (prog && prog->aux->sleepable)
 			return -EINVAL;
 	}
 
Signature · 1 line
2.34.1
#2 lore raw

PATCH v4 2/3 bpf: Check cfi_stubs before registering a struct_ops type.

In reply to [email protected]

From: Kui-Feng Lee <[email protected]>

Recently, cfi_stubs were introduced. However, existing struct_ops types that are not in the upstream may not be aware of this, resulting in kernel crashes. By rejecting struct_ops types that do not provide cfi_stubs during registration, these crashes can be avoided.

---
 kernel/bpf/bpf_struct_ops.c | 17 +++++++++++++++++
 1 file changed, 17 insertions(+)
kernel/bpf/bpf_struct_ops.c +17 -0
index 0d7be97a2411..c1c502caae08 100644
@@ -302,6 +302,11 @@ int bpf_struct_ops_desc_init(struct bpf_struct_ops_desc *st_ops_desc,
 	}
 	sprintf(value_name, "%s%s", VALUE_PREFIX, st_ops->name);
 
+	if (!st_ops->cfi_stubs) {
+		pr_warn("struct %s has no cfi_stubs\n", st_ops->name);
+		return -EINVAL;
+	}
+
 	type_id = btf_find_by_name_kind(btf, st_ops->name,
 					BTF_KIND_STRUCT);
 	if (type_id < 0) {
@@ -339,6 +344,7 @@ int bpf_struct_ops_desc_init(struct bpf_struct_ops_desc *st_ops_desc,
 
 	for_each_member(i, t, member) {
 		const struct btf_type *func_proto;
+		u32 moff;
 
 		mname = btf_name_by_offset(btf, member->name_off);
 		if (!*mname) {
@@ -361,6 +367,17 @@ int bpf_struct_ops_desc_init(struct bpf_struct_ops_desc *st_ops_desc,
 		if (!func_proto)
 			continue;
 
+		moff = __btf_member_bit_offset(t, member) / 8;
+		err = st_ops->check_member ?
+			st_ops->check_member(t, member, NULL) : 0;
+
+		if (!err && !*(void **)(st_ops->cfi_stubs + moff)) {
+			pr_warn("member %s in struct %s has no cfi stub function\n",
+				mname, st_ops->name);
+			err = -EINVAL;
+			goto errout;
+		}
+
 		if (btf_distill_func_proto(log, btf,
 					   func_proto, mname,
 					   &st_ops->func_models[i])) {
Signature · 1 line
2.34.1
#6 lore raw

PATCH v4 3/3 selftests/bpf: Test case for lacking CFI stub functions.

In reply to [email protected]

From: Kui-Feng Lee <[email protected]>

Ensure struct_ops rejects the registration of struct_ops types without proper CFI stub functions.

bpf_test_no_cfi.ko is a module that attempts to register a struct_ops type called "bpf_test_no_cfi_ops" with varying levels of cfi_stubs. It starts with a NULL cfi_stub and ends with a fully complete cfi_stub. Only the fully complete cfi_stub should be accepted by struct_ops. The module can only be loaded successfully if these registrations yield the expected results.

---
 tools/testing/selftests/bpf/Makefile          | 10 +-
 .../selftests/bpf/bpf_test_no_cfi/Makefile    | 19 ++++
 .../bpf/bpf_test_no_cfi/bpf_test_no_cfi.c     | 93 +++++++++++++++++++
 .../bpf/prog_tests/test_struct_ops_no_cfi.c   | 38 ++++++++
 tools/testing/selftests/bpf/testing_helpers.c |  4 +-
 tools/testing/selftests/bpf/testing_helpers.h |  2 +
 6 files changed, 163 insertions(+), 3 deletions(-)
 create mode 100644 tools/testing/selftests/bpf/bpf_test_no_cfi/Makefile
 create mode 100644 tools/testing/selftests/bpf/bpf_test_no_cfi/bpf_test_no_cfi.c
 create mode 100644 tools/testing/selftests/bpf/prog_tests/test_struct_ops_no_cfi.c
tools/testing/selftests/bpf/Makefile +9 -1
index dbb8c5f94f34..c219da5e60e6 100644
@@ -132,7 +132,7 @@ TEST_GEN_PROGS_EXTENDED = test_sock_addr test_skb_cgroup_id_user \
 	flow_dissector_load test_flow_dissector test_tcp_check_syncookie_user \
 	test_lirc_mode2_user xdping test_cpp runqslower bench bpf_testmod.ko \
 	xskxceiver xdp_redirect_multi xdp_synproxy veristat xdp_hw_metadata \
-	xdp_features
+	xdp_features bpf_test_no_cfi.ko
 
 TEST_GEN_FILES += liburandom_read.so urandom_read sign-file uprobe_multi
 
@@ -254,6 +254,12 @@ $(OUTPUT)/bpf_testmod.ko: $(VMLINUX_BTF) $(RESOLVE_BTFIDS) $(wildcard bpf_testmo
 	$(Q)$(MAKE) $(submake_extras) RESOLVE_BTFIDS=$(RESOLVE_BTFIDS) -C bpf_testmod
 	$(Q)cp bpf_testmod/bpf_testmod.ko $@
 
+$(OUTPUT)/bpf_test_no_cfi.ko: $(VMLINUX_BTF) $(RESOLVE_BTFIDS) $(wildcard bpf_test_no_cfi/Makefile bpf_test_no_cfi/*.[ch])
+	$(call msg,MOD,,$@)
+	$(Q)$(RM) bpf_test_no_cfi/bpf_test_no_cfi.ko # force re-compilation
+	$(Q)$(MAKE) $(submake_extras) RESOLVE_BTFIDS=$(RESOLVE_BTFIDS) -C bpf_test_no_cfi
+	$(Q)cp bpf_test_no_cfi/bpf_test_no_cfi.ko $@
+
 DEFAULT_BPFTOOL := $(HOST_SCRATCH_DIR)/sbin/bpftool
 ifneq ($(CROSS_COMPILE),)
 CROSS_BPFTOOL := $(SCRATCH_DIR)/sbin/bpftool
@@ -628,6 +634,7 @@ TRUNNER_EXTRA_SOURCES := test_progs.c		\
 			 flow_dissector_load.h	\
 			 ip_check_defrag_frags.h
 TRUNNER_EXTRA_FILES := $(OUTPUT)/urandom_read $(OUTPUT)/bpf_testmod.ko	\
+		       $(OUTPUT)/bpf_test_no_cfi.ko			\
 		       $(OUTPUT)/liburandom_read.so			\
 		       $(OUTPUT)/xdp_synproxy				\
 		       $(OUTPUT)/sign-file				\
@@ -756,6 +763,7 @@ EXTRA_CLEAN := $(SCRATCH_DIR) $(HOST_SCRATCH_DIR)			\
 	feature bpftool							\
 	$(addprefix $(OUTPUT)/,*.o *.skel.h *.lskel.h *.subskel.h	\
 			       no_alu32 cpuv4 bpf_gcc bpf_testmod.ko	\
+			       bpf_test_no_cfi.ko			\
 			       liburandom_read.so)
 
 .PHONY: docs docs-clean
tools/testing/selftests/bpf/bpf_test_no_cfi/Makefile new file +19 -0
new file mode 100644
index 000000000000..ed5143b79edf
@@ -0,0 +1,19 @@
+BPF_TEST_NO_CFI_DIR := $(realpath $(dir $(abspath $(lastword $(MAKEFILE_LIST)))))
+KDIR ?= $(abspath $(BPF_TEST_NO_CFI_DIR)/../../../../..)
+
+ifeq ($(V),1)
+Q =
+else
+Q = @
+endif
+
+MODULES = bpf_test_no_cfi.ko
+
+obj-m += bpf_test_no_cfi.o
+
+all:
+	+$(Q)make -C $(KDIR) M=$(BPF_TEST_NO_CFI_DIR) modules
+
+clean:
+	+$(Q)make -C $(KDIR) M=$(BPF_TEST_NO_CFI_DIR) clean
+
tools/testing/selftests/bpf/bpf_test_no_cfi/bpf_test_no_cfi.c new file +93 -0

93 lines not on this page · raw message

tools/testing/selftests/bpf/prog_tests/test_struct_ops_no_cfi.c new file +38 -0
new file mode 100644
index 000000000000..19703e250549
@@ -0,0 +1,38 @@
+// SPDX-License-Identifier: GPL-2.0
+/* Copyright (c) 2024 Meta Platforms, Inc. and affiliates. */
+#include <test_progs.h>
+#include <testing_helpers.h>
+
+static void load_bpf_test_no_cfi(void)
+{
+	int fd;
+	int err;
+
+	fd = open("bpf_test_no_cfi.ko", O_RDONLY);
+	if (!ASSERT_GT(fd, 0, "open")) {
+		close(fd);
+		return;
+	}
+
+	/* The module will try to register a struct_ops type with
+	 * no cfi_stubs, incomplete cfi_stubs, and full cfi_stubs.
+	 *
+	 * Only full cfi_stubs should be allowed. The module will be loaded
+	 * successfully if the result of the registration is as expected,
+	 * or it fails.
+	 */
+	err = finit_module(fd, "", 0);
+	close(fd);
+	if (!ASSERT_OK(err, "finit_module"))
+		return;
+
+	err = delete_module("bpf_test_no_cfi", 0);
+	ASSERT_OK(err, "delete_module");
+}
+
+void test_struct_ops_no_cfi(void)
+{
+	if (test__start_subtest("load_bpf_test_no_cfi"))
+		load_bpf_test_no_cfi();
+}
+
tools/testing/selftests/bpf/testing_helpers.c +2 -2
index a59e56d804ee..28b6646662af 100644
@@ -356,12 +356,12 @@ __u64 read_perf_max_sample_freq(void)
 	return sample_freq;
 }
 
-static int finit_module(int fd, const char *param_values, int flags)
+int finit_module(int fd, const char *param_values, int flags)
 {
 	return syscall(__NR_finit_module, fd, param_values, flags);
 }
 
-static int delete_module(const char *name, int flags)
+int delete_module(const char *name, int flags)
 {
 	return syscall(__NR_delete_module, name, flags);
 }
tools/testing/selftests/bpf/testing_helpers.h +2 -0
index d14de81727e6..d55f6ab12433 100644
@@ -36,6 +36,8 @@ __u64 read_perf_max_sample_freq(void);
 int load_bpf_testmod(bool verbose);
 int unload_bpf_testmod(bool verbose);
 int kern_sync_rcu(void);
+int finit_module(int fd, const char *param_values, int flags);
+int delete_module(const char *name, int flags);
 
 static inline __u64 get_time_ns(void)
 {
Signature · 1 line
2.34.1
#3 lore raw

Re: [PATCH bpf-next v4 2/3] bpf: Check cfi_stubs before registering a struct_ops type.

In reply to [email protected]
Quoted 37 lines · On 2/20/24 11:52 PM, [email protected] wrote:

From: Kui-Feng Lee <[email protected]>

Recently, cfi_stubs were introduced. However, existing struct_ops types that are not in the upstream may not be aware of this, resulting in kernel crashes. By rejecting struct_ops types that do not provide cfi_stubs during registration, these crashes can be avoided.

---
  kernel/bpf/bpf_struct_ops.c | 17 +++++++++++++++++
  1 file changed, 17 insertions(+)
kernel/bpf/bpf_struct_ops.c +6 -0
index 0d7be97a2411..c1c502caae08 100644
@@ -302,6 +302,11 @@ int bpf_struct_ops_desc_init(struct bpf_struct_ops_desc *st_ops_desc,
  	}
  	sprintf(value_name, "%s%s", VALUE_PREFIX, st_ops->name);
 
+	if (!st_ops->cfi_stubs) {
+		pr_warn("struct %s has no cfi_stubs\n", st_ops->name);
+		return -EINVAL;
+	}
+
  	type_id = btf_find_by_name_kind(btf, st_ops->name,
  					BTF_KIND_STRUCT);
  	if (type_id < 0) {
@@ -339,6 +344,7 @@ int bpf_struct_ops_desc_init(struct bpf_struct_ops_desc *st_ops_desc,
 
  	for_each_member(i, t, member) {
  		const struct btf_type *func_proto;
+		u32 moff;
 
  		mname = btf_name_by_offset(btf, member->name_off);
  		if (!*mname) {

kernel/bpf/bpf_struct_ops.c int bpf_struct_ops_desc_init(struct bpf_struct_ops_desc *st_ops_desc, · line 372

Quoted 6 lines
kernel/bpf/bpf_struct_ops.c +3 -0
@@ -361,6 +367,17 @@ int bpf_struct_ops_desc_init(struct bpf_struct_ops_desc *st_ops_desc,
  		if (!func_proto)
  			continue;
 
+		moff = __btf_member_bit_offset(t, member) / 8;
+		err = st_ops->check_member ?
+			st_ops->check_member(t, member, NULL) : 0;

I don't think it is necessary to make check_member more complicated by taking NULL prog. The struct_ops implementer then needs to handle this extra NULL prog case.

Have you thought about Alexei's earlier suggestion in v3 to reuse the NULL member in cfi_stubs to flag unsupported member and remove the unsupported_ops[] from bpf_tcp_ca.c?

If the verifier can consistently reject loading unsupported bpf prog, it will not reach the bpf_struct_ops_map_update_elem and then hits the NULL member in cfi_stubs during map_update_elem.

Untested code:

kernel/bpf/verifier.c +9 -1
index 011d54a1dc53..c57cb0e2a8a7 100644
@@ -20370,6 +20370,7 @@ static int check_struct_ops_btf_id(struct bpf_verifier_env *env)
 	u32 btf_id, member_idx;
 	struct btf *btf;
 	const char *mname;
+	u32 moff;
 
 	if (!prog->gpl_compatible) {
 		verbose(env, "struct ops programs must have a GPL compatible license\n");
@@ -20417,11 +20418,18 @@ static int check_struct_ops_btf_id(struct bpf_verifier_env *env)
 		return -EINVAL;
 	}
 
+	moff = __btf_member_bit_offset(t, member) / 8;
+	if (!*(void **)(st_ops->cfi_stubs + moff)) {
+		verbose(env, "attach to unsupported member %s of struct %s\n",
+			mname, st_ops->name);
+		return -ENOTSUPP;
+	}
+
 	if (st_ops->check_member) {
 		int err = st_ops->check_member(t, member, prog);
 
 		if (err) {
-			verbose(env, "attach to unsupported member %s of struct %s\n",
+			verbose(env, "cannot attach to member %s of struct %s\n",
 				mname, st_ops->name);
 			return err;
 		}
net/ipv4/bpf_tcp_ca.c +0 -26
index 7f518ea5f4ac..bcb1fcd00973 100644
@@ -14,10 +14,6 @@
 /* "extern" is to avoid sparse warning.  It is only used in bpf_struct_ops.c. */
 static struct bpf_struct_ops bpf_tcp_congestion_ops;
 
-static u32 unsupported_ops[] = {
-	offsetof(struct tcp_congestion_ops, get_info),
-};
-
 static const struct btf_type *tcp_sock_type;
 static u32 tcp_sock_id, sock_id;
 static const struct btf_type *tcp_congestion_ops_type;
@@ -45,18 +41,6 @@ static int bpf_tcp_ca_init(struct btf *btf)
 	return 0;
 }
 
-static bool is_unsupported(u32 member_offset)
-{
-	unsigned int i;
-
-	for (i = 0; i < ARRAY_SIZE(unsupported_ops); i++) {
-		if (member_offset == unsupported_ops[i])
-			return true;
-	}
-
-	return false;
-}
-
 static bool bpf_tcp_ca_is_valid_access(int off, int size,
 				       enum bpf_access_type type,
 				       const struct bpf_prog *prog,
@@ -248,15 +232,6 @@ static int bpf_tcp_ca_init_member(const struct btf_type *t,
 	return 0;
 }
 
-static int bpf_tcp_ca_check_member(const struct btf_type *t,
-				   const struct btf_member *member,
-				   const struct bpf_prog *prog)
-{
-	if (is_unsupported(__btf_member_bit_offset(t, member) / 8))
-		return -ENOTSUPP;
-	return 0;
-}
-
 static int bpf_tcp_ca_reg(void *kdata)
 {
 	return tcp_register_congestion_control(kdata);
@@ -350,7 +325,6 @@ static struct bpf_struct_ops bpf_tcp_congestion_ops = {
 	.reg = bpf_tcp_ca_reg,
 	.unreg = bpf_tcp_ca_unreg,
 	.update = bpf_tcp_ca_update,
-	.check_member = bpf_tcp_ca_check_member,
 	.init_member = bpf_tcp_ca_init_member,
 	.init = bpf_tcp_ca_init,
 	.validate = bpf_tcp_ca_validate,
Signature · 1 line
2.34.1
Quoted 11 lines
kernel/bpf/bpf_struct_ops.c +8 -0
+
+		if (!err && !*(void **)(st_ops->cfi_stubs + moff)) {
+			pr_warn("member %s in struct %s has no cfi stub function\n",
+				mname, st_ops->name);
+			err = -EINVAL;
+			goto errout;
+		}
+
  		if (btf_distill_func_proto(log, btf,
  					   func_proto, mname,
  					   &st_ops->func_models[i])) {
#4 lore raw

Re: [PATCH bpf-next v4 2/3] bpf: Check cfi_stubs before registering a struct_ops type.

Quoted 48 lines · Martin KaFai Lau
Quoted 36 lines · On 2/20/24 11:52 PM, [email protected] wrote:

From: Kui-Feng Lee <[email protected]>

Recently, cfi_stubs were introduced. However, existing struct_ops types that are not in the upstream may not be aware of this, resulting in kernel crashes. By rejecting struct_ops types that do not provide cfi_stubs during registration, these crashes can be avoided.

---
  kernel/bpf/bpf_struct_ops.c | 17 +++++++++++++++++

  1 file changed, 17 insertions(+)
kernel/bpf/bpf_struct_ops.c +6 -0
index 0d7be97a2411..c1c502caae08 100644
@@ -302,6 +302,11 @@ int bpf_struct_ops_desc_init(struct 
 bpf_struct_ops_desc *st_ops_desc,
       }
       sprintf(value_name, "%s%s", VALUE_PREFIX, st_ops->name);
+    if (!st_ops->cfi_stubs) {
+        pr_warn("struct %s has no cfi_stubs\n", st_ops->name);
+        return -EINVAL;
+    }
+
       type_id = btf_find_by_name_kind(btf, st_ops->name,
                       BTF_KIND_STRUCT);
       if (type_id < 0) {
@@ -339,6 +344,7 @@ int bpf_struct_ops_desc_init(struct 
 bpf_struct_ops_desc *st_ops_desc,
       for_each_member(i, t, member) {
           const struct btf_type *func_proto;
+        u32 moff;
           mname = btf_name_by_offset(btf, member->name_off);
           if (!*mname) {

kernel/bpf/bpf_struct_ops.c int bpf_struct_ops_desc_init(struct · line 372

Quoted 6 lines
kernel/bpf/bpf_struct_ops.c +3 -0
@@ -361,6 +367,17 @@ int bpf_struct_ops_desc_init(struct 
 bpf_struct_ops_desc *st_ops_desc,
           if (!func_proto)
               continue;
+        moff = __btf_member_bit_offset(t, member) / 8;
+        err = st_ops->check_member ?
+            st_ops->check_member(t, member, NULL) : 0;

I don't think it is necessary to make check_member more complicated by taking NULL prog. The struct_ops implementer then needs to handle this extra NULL prog case.

Quoted 8 lines

Have you thought about Alexei's earlier suggestion in v3 to reuse the NULL member in cfi_stubs to flag unsupported member and remove the unsupported_ops[] from bpf_tcp_ca.c?

If the verifier can consistently reject loading unsupported bpf prog, it will not reach the bpf_struct_ops_map_update_elem and then hits the NULL member in cfi_stubs during map_update_elem.

Ok! I misunderstood previously. I will go this way.

#5 lore raw

Re: [PATCH bpf-next v4 2/3] bpf: Check cfi_stubs before registering a struct_ops type.

Quoted 55 lines · Kui-Feng Lee
Quoted 52 lines · Martin KaFai Lau
Quoted 36 lines · On 2/20/24 11:52 PM, [email protected] wrote:

From: Kui-Feng Lee <[email protected]>

Recently, cfi_stubs were introduced. However, existing struct_ops types that are not in the upstream may not be aware of this, resulting in kernel crashes. By rejecting struct_ops types that do not provide cfi_stubs during registration, these crashes can be avoided.

---
  kernel/bpf/bpf_struct_ops.c | 17 +++++++++++++++++

  1 file changed, 17 insertions(+)
kernel/bpf/bpf_struct_ops.c +6 -0
index 0d7be97a2411..c1c502caae08 100644
@@ -302,6 +302,11 @@ int bpf_struct_ops_desc_init(struct 
 bpf_struct_ops_desc *st_ops_desc,
       }
       sprintf(value_name, "%s%s", VALUE_PREFIX, st_ops->name);
+    if (!st_ops->cfi_stubs) {
+        pr_warn("struct %s has no cfi_stubs\n", st_ops->name);
+        return -EINVAL;
+    }
+
       type_id = btf_find_by_name_kind(btf, st_ops->name,
                       BTF_KIND_STRUCT);
       if (type_id < 0) {
@@ -339,6 +344,7 @@ int bpf_struct_ops_desc_init(struct 
 bpf_struct_ops_desc *st_ops_desc,
       for_each_member(i, t, member) {
           const struct btf_type *func_proto;
+        u32 moff;
           mname = btf_name_by_offset(btf, member->name_off);
           if (!*mname) {

kernel/bpf/bpf_struct_ops.c int bpf_struct_ops_desc_init(struct · line 372

Quoted 6 lines
kernel/bpf/bpf_struct_ops.c +3 -0
@@ -361,6 +367,17 @@ int bpf_struct_ops_desc_init(struct 
 bpf_struct_ops_desc *st_ops_desc,
           if (!func_proto)
               continue;
+        moff = __btf_member_bit_offset(t, member) / 8;
+        err = st_ops->check_member ?
+            st_ops->check_member(t, member, NULL) : 0;

I don't think it is necessary to make check_member more complicated by taking NULL prog. The struct_ops implementer then needs to handle this extra NULL prog case.

Have you thought about Alexei's earlier suggestion in v3 to reuse the NULL member in cfi_stubs to flag unsupported member and remove the unsupported_ops[] from bpf_tcp_ca.c?

Quoted 7 lines
Quoted 4 lines

If the verifier can consistently reject loading unsupported bpf prog, it will not reach the bpf_struct_ops_map_update_elem and then hits the NULL member in cfi_stubs during map_update_elem.

Ok! I misunderstood previously. I will go this way.

According to the off-line discussion, the changes for unsupported_ops[] should be in a separate patchset. The check of (void **)(st_ops->cfi_stubs + moff)) will be removed. Changes of check_member should be removed as well.